Exposure apparatus and method of manufacturing device

1. Field of the Invention

The present invention relates to an exposure apparatus and a method of manufacturing a device and, more particularly, to an exposure apparatus used in manufacturing a device such as a semiconductor device, an image sensing device, a liquid crystal device, or a thin-film magnetic head by lithography, and a method of manufacturing a device using the same.

2. Description of the Related Art

Recently, in the quest for smaller, thinner electronic devices, demands for advances in micropatterning of semiconductor devices mounted in the electronic devices are further growing. For example, the design rule is about to achieve the formation of a circuit pattern having a minimum feature size of 100 nm or less on mass production lines, and is expected to further shift to the formation of a circuit pattern having a minimum feature size of 80 nm or less in the future. A main stream processing technique for attaining such a strict design rule is photolithography. In photolithography, a projection exposure apparatus which projects and transfers a mask pattern drawn on a mask or reticle onto a wafer using a projection optical system has conventionally been employed.

A resolution R of the projection exposure apparatus is given by the Rayleigh equation:

R 32 k₁×λ/NA   (1)

where λ is the wavelength of the light source, and NA is the numerical aperture of the projection optical system.

A depth of focus DOF, that is, the focal range within which a constant image forming performance can be maintained is given by:

DOF=k₂×λ/NA²   (2)

As the DOF decreases, focusing becomes more difficult. To overcome this difficulty, there is a demand to increase the substrate flatness and the focusing accuracy, and so, essentially, the DOF is desirably large.

The mask pattern includes, for example, a line & space (L & S) pattern having adjacent lines and spaces, a contact hole array having adjacent and periodical contact holes (i.e., contact holes arrayed at an interval equal to the hole diameter), an isolated contact hole which is not adjacent but isolated, and other isolated patterns. However, to transfer the pattern with a high resolution, it is necessary to select an optimum illumination condition in accordance with the type of pattern.

System chips manufactured by the modern semiconductor industry are shifting to those having a higher added value and a mixture of a wide variety of patterns. Under the circumstances, it has become necessary to fabricate a mask having a mixture of a plurality of types of contact hole patterns. However, it has been impossible to simultaneously transfer contact hole patterns each having a mixture of a contact hole array and an isolated contact hole with a high resolution.

To combat this situation, various kinds of methods have been proposed for increasing the depth of focus by increasing the resolution limits of only a contact hole array and a two-dimensional repetitive interconnection pattern. An example of these methods is a phase shift method using a double exposure(or multiple exposure) scheme which separately forms by exposure different types of patterns using two masks, or a scheme which enhances the resolving power of a main pattern by providing various types of auxiliary patterns to the mask pattern. This method improves the resolving power by forming a thin film which guides the propagating light so that a part of the conventional mask becomes 180° out of phase with its remaining part.

Unfortunately, various problems remain to be solved in order to improve the resolving power using a phase shift mask of a type which actually modulates the spatial frequency. Because of these problems, it is currently very difficult to manufacture semiconductor devices by actually using the phase shift mask.

A method commonly used at present exposes one mask under a special illumination condition. In contrast to vertical illumination as conventional illumination, this method obliquely applies light onto the reticle by adjusting the effective light source shape to an annular shape or a quadrupole shape, and is called modified illumination (off-axis illumination).

In conventional illumination, an image is formed by interference among three light beams: the 0th- and ±1st-order light beams. In this case, the ±1st-order light beams are distributed at positions, which are shifted by their diffraction angles from the optical axis, in the pupil plane, as shown in FIG. 4A. However, as shown in FIG. 4B, as the pattern becomes finer, the intervals between the 0-th and ±1st-order light beams widen, so a certain component of the diffracted light falls outside the aperture stop of the projection lens. As a consequence, the 0th-order light beam can interfere with no light beam, and therefore no image can be formed.